Channel quality prediction and localization

Abstract

In this thesis we focus on two problems, Channel Quality Prediction and Localization in wireless network. In High-Speed Downlink Packet Access (HSDPA) architecture of Universal Mobile Telecommunications Services (UMTS), User Equipment (UE) provides Channel Quality Indicator (CQI) values to Node B (base station), as an indication of current channel quality. We investigate two methods for channel quality prediction, which can be implemented at Node B. Each of these methods uses recent CQI values to predict future channel quality, i.e. future CQI values. Experiments are conducted in different environments, to check accuracy of these methods and found that the simple method using Least-Squares Line gives as good results as the complex method using Spline. Packet schedulers can be developed which uses predicted CQI values for scheduling. Cellular networks are used for localization where the UEs are located by measuring the signal traveling to and from a set of fixed cellular base stations (Node B). We propose a Markov Chain Model, which can be used as supplement to Angle of Arrival (AOA) method which is used for localization. Sector-specific information with recent CQI values and information about angle of arrival of signal can be used to estimate UE’s location using proposed model, in UMTS network supporting HSDPA architecture.